Why 18 Months Makes the Difference.

At some point, usually around the third redesign cycle, someone leans back in their chair and says it:

“Why is this suddenly so expensive?”

Good question. Familiar one, too.

We’ve spent decades picking up the pieces of laboratories that went sideways on cost. Wet labs, GMP suites, cleanrooms with loyalty to five different regulators at once. And if there’s one pattern we see again and again, it’s this:

Engineering joins the party far too late. We’re talking structurally, financially, painfully late.

The £5m decision that didn’t look like a decision

Here’s the thing nobody tells you early enough: Most lab cost blowouts don’t happen during construction. They don’t even happen during design development. They happen earlier. When nobody is measuring.

They happen when floor-to-floor heights are “probably fine.” When plant is assumed to “fit later.” When services are imagined as lines on drawings, not solid objects with opinions and spatial demands.

This is where engineering often becomes reactive, asked to ‘make it work’ inside a box that was already sized, shaped, and priced based on assumptions. Optimistic ones.

Labs are not offices with sinks

Somewhere along the line, someone decided a lab is basically an office building with extra plumbing. This belief has cost the industry billions.

Laboratories are energy-hungry, vibration-sensitive, regulation-soaked beasts. They care deeply about air changes, pressure regimes, redundancy, maintenance access, commissioning strategy, and what happens after handover, when the science starts.

You can’t bolt that on at the end. You can try. People do. Then ceilings drop. Slabs thicken. Cost follows. Relentlessly.

So why 18 months?

Because that’s roughly how long it takes for a lab project to move from “idea” to “irreversible.”

By 18 months before construction, the big decisions are still fluid:

• Structural grid and slab thickness

• Floor-to-floor heights

• Plant strategy

• Utility capacity vs. real demand

• Flexibility for unknown future tenants

• CQV approach as a design driver

Miss that window, and you’re compensating. And compensation is expensive.

Early engineering is about direction

People think early engineering means specifying equipment models at feasibility. Or drowning a concept in technical diagrams. No. Absolutely not.

Early engineering is about asking uncomfortable questions before the answers get costly:

• What kind of science could realistically happen here in five years?

• What regulatory ceiling are we designing for, even if we don’t hit it on day one?

• Where are the non-negotiables hiding?

• Which systems must be future-proof, and which can be value-engineered safely?

It’s about avoiding blind alleys.

The GMP trap

GMP is a favourite culprit.

Projects often start with a vague intention:
“Maybe GMP-ready. Not full GMP. Yet.”

Sounds sensible. Capital-light. Flexible.

Until someone realises, too late, that slab heights, pressure regimes, maintenance access, zoning, and validation pathways were never aligned with any credible GMP trajectory.

Suddenly “GMP-ready” means ripping things out. Or worse, building elaborate workarounds that satisfy auditors but terrify cost consultants. If engineering had been involved 18 months earlier, the conversation would have been calmer. Smarter. Cheaper.

Value engineering vs. value erosion

Another uncomfortable truth: late engineering erodes value. When budgets tighten late, things get cut. Usually the wrong things. Redundancy goes. Flexibility shrinks. Commissioning time compresses (always a bad idea, by the way). The building opens fragile. It works, but only just.

Six months later, operational costs creep up. Then maintenance pain. Then tenant dissatisfaction. Then, inevitably, retrofit.

A success story

We worked on a lab project where engineering was invited in before the site was even purchased.

We spent months doing things that looked slow from the outside: modelling scenarios, stress-testing assumptions, debating over ceiling zones, sketching plant layouts long before architects were worrying about finishes.

No heroics later.
No panicked redesign.
No emergency overruns “nobody could have predicted.”

It wasn’t flashy. It just… worked.

Those are our favourite projects. Methodology driven projects.

What this means for developers, investors, and operators

If you’re an investor: early engineering protects downside risk. Full stop.

If you’re a developer: it preserves optionality, the kind you can actually monetise.

If you’re an operator: it saves you from inheriting someone else’s shortcuts.

And if you’re a scientist? It means a building that supports your work instead of fighting it.

Some uncomfortable questions worth asking (now, not later)

• Are we designing for the science we want, or the science this building can honestly support?

• What assumptions are hidden in our cost plan?

• What happens if a tenant needs more air, more power, or stricter controls than we expected?

• Who’s thinking about commissioning, validation, and operational readiness, and when?

If these questions feel awkward, good. They should. That’s the moment where cost certainty is still possible.

Final thought

A handful of lab projects fail because ambition and engineering didn’t meet early enough to have an honest conversation.

Eighteen months is just often the last sensible moment to choose strategy over salvage.

So maybe the better question isn’t,
“Can we afford to bring engineering in earlier?”

It’s:
“What does it really cost when we don’t?”

Let’s talk about that.